Commit Graph

91 Commits

Author SHA1 Message Date
tg 328cbf510c Unicode 6.3.0 2013-11-30 23:20:04 +00:00
tg 3bc3e8665a fix postun in not evaluated side of e.g. ternary operator (LP#1187729) 2013-07-21 18:38:56 +00:00
tg 49d1ebfb71 Replace wcwidth code by mine based on Unicode 6.2.0 2013-05-31 23:27:14 +00:00
tg 75a4809a3a fix remaining signed int nonsense I am aware of 2013-04-26 19:10:58 +00:00
tg 1df3efdb86 another intermediate step/commit to get arithmetics right:
do the correct operations for comparisons (just keep using the
signed/unsigned switch from bivui for them), division (by working
on absolutes and adding the sign at the end), modulo (stupidly by
divising in signed, multiplying and subtracting, to get the sign
of the result right)

also adds rotation

XXX to check: do we need to AND before assigning the result in division?
2013-04-14 13:36:53 +00:00
tg 7cce9de0bc one can’t cast an unsigned int to a signed int portably either, if the value
is larger than the positive range of the latter (implementation-defined), so
avoid them in all explicit cases and rearrange stuff and check for it

(I’m gonna have to revise lots more code…)
2013-04-01 02:37:53 +00:00
tg fb3555d7f5 got an SAR, thanks to dalias on IRC 2013-04-01 02:28:35 +00:00
tg e1a4c3ed28 do use val.u on the RHS of a shift 2013-04-01 01:29:47 +00:00
tg 0f417614dd ugh, signed >> on negative values is implementation-defined
(but, at least, not undefined, and usually right; regress-test for it)
2013-04-01 01:16:37 +00:00
tg d96c733069 looks like multiplication is also the same between signed and unsigned 2013-04-01 01:02:09 +00:00
tg 6a745a01fa begin using unsigned arithmetics internally as much as possible
(i.e. where signedness doesn’t matter, given -fwrapv) and note
where there’s work to do; note future improvements and additions;
optimise a little

mostly untested
2013-03-31 18:33:13 +00:00
tg 24b81b5969 RCSID-only sync: oksh found out about idivwrapv 2013-03-29 16:54:05 +00:00
tg 12f0ff60d2 prevent more access to invalid memories (Debian #700604) 2013-02-15 18:36:48 +00:00
tg d8662eb228 even more hacks to pass “-O666 -fstrict-overflow -Wstrict-overflow=9 -flto=jobserver” with “gcc version 4.8.0 20120930 (experimental) [trunk revision 191865] (Debian 20120930-1)” ☹ plus make the rtchecks mandatory 2012-10-03 17:24:23 +00:00
tg 0415b74436 Update wcwidth data from Unicode 6.1.0 2012-09-01 23:46:41 +00:00
tg c1f821d4e5 (mksh) tighten 32-bit requirements; (lksh) switch to long; allow any bitness 2012-06-28 20:17:39 +00:00
tg dc5ae267ce make tempvar() and vtemp global 2012-06-28 20:02:29 +00:00
tg 9646c98356 I give up on writing (1 << 31) in any form.
Now we just use 0x80000000UL and cast that to mksh_uari_t and,
if required, that to mksh_ari_t afterwards.
2012-03-31 17:52:33 +00:00
tg e67b98e21b use _setjmp/_longjmp on NeXTstep: its Intel port always restores the signal mask on siglongjmp, which we never have 2012-03-31 17:30:00 +00:00
tg cf75e7b6ce couple of minor/cosmetic fixes from RT’s compile farm:
• promote SCO OpenServer and UnixWare to !oswarn
• omit trying -O2/-O on OpenServer 5 and USL C
• cast mksh_ari_t to int, mksh_uari_t to unsigned int for printf
• skip ulimit-1 on syllable (which is still too broken)
• write ((mksh_ari_t)-2147483648) ipv UB ((mksh_ari_t)1 << 31)
  and add a comment that that is actually meant
• rewrite functions returning !void ending in NOTREACHED
  so they’ve got a jump target returning an error at the
  end, to aid older compilers and just to be safe
• cast struct stat.st_size to off_t or size_t explicitly when needed
• shorten struct env by two bytes and an alignment, at least

also, optimise control flow and fix more paren matching cases
2012-03-29 19:23:01 +00:00
tg 7ec1ae3771 followup for cid 1004EE408E1382C1752 and 1004EE40DDD498FBB0D:
do a mirtoconf run-time check (ugh) to see whether the CPU designers
smoked/were brain-dead or if we don’t actually need the manual check
2011-12-31 02:04:18 +00:00
tg 5070ea2387 skip R/O check in an unevaluated ternary part; 10x jilles 2011-12-16 20:03:02 +00:00
tg e18f4d114a actually, behave with silent wraparound; results validated by bc(1) 2011-12-11 01:56:43 +00:00
tg 971b153933 catch intmin/-1 instead of dumping core on SIGFPE; from Jilles Tjoelker 2011-12-11 01:35:10 +00:00
tg 9782f6b4d1 • access(2) is broken in at least kFreeBSD 9.0 as “modern” OS, so bring
back the wrapper code as well as refactor most other code calling it
• apparently, names can’t end in ‘_’ or contain ‘__’ anywhere…
2011-09-07 15:24:22 +00:00
tg 577c918beb patch most of Jerker Bäck’s concerns out, unless not applicable 2011-08-27 18:06:52 +00:00
tg 210608b199 regenerate all wcwidth code from UCD 6.0.0 and related tables,
and mgk25’s wcwidth.c 5.0 code; add check against 2-byte wchar_t;
optimise libc wcwidth(3) implementation taken from mksh and sync these
2010-12-11 16:05:03 +00:00
tg e57379aab7 regenerate from Unicode 6.0.0 2010-11-01 17:28:49 +00:00
tg 3747722db4 improve string pooling: saves 316 bytes in .text 2010-08-28 18:50:58 +00:00
tg 23f3f58d14 on MirBSD we can use the system wcwidth() and save ~800by on the ramdisc 2010-08-14 21:35:13 +00:00
tg f3b3b4b1fb remove some debugging code 2010-01-25 14:38:04 +00:00
tg 769e222586 re-vamp __attribute__ handling; let this pass on HP-UX bundled compiler
as well as HP aCC
2009-12-12 22:27:10 +00:00
tg a77bbf6123 HP aCC tells me I'm using "const" twice, and /usr/ccs/bin/cc even
errors out here. Let's hope that this form will *still* have the
entirety of that array in .rodata...

XXX more to come, thanks to HP DSPP PvP, e.g. __attribute__ stuff
2009-12-08 19:23:34 +00:00
tg 703551bd4f attempt a better fix: break off widthadj upon encountering NUL,
add remaining columns as octets (should also speed optimise)
2009-12-05 20:17:59 +00:00
tg a09f05e77a /me is annoyed
I read, IIRC in the Cederqvist, that 'cvs tag' sets a sticky tag onto
the cwd… it doesn’t, apparently. (I actually like it better this way,
but one needs to know!)
2009-11-28 14:28:03 +00:00
tg 883d9d99b3 switch ${%foo} to wcswidth-like behaviour – slightly problematic, and
the “set +U” case isn’t even handled

committed to branch because I’d like to get more input on this, for now
2009-11-28 14:21:47 +00:00
tg 73147a7fe9 fix lazy evaluation of assignments in ternary ops 2009-10-04 13:19:33 +00:00
tg 7b7b75b026 * move the utf_* functions to a smaller file, to reduce the pain the
CPU has to endure while gcc is crunching on edit.c
* comment on mksh not using _exactly_ OPTU-8/OPTU-16 (XXX)
2009-09-26 04:01:34 +00:00
tg 1a28786229 * shrink MKSH_SMALL even further by removing functionality like
some GNU bash extensions (suggested by cnuke@) and bind macros
* make the random cache more efficient (and the code potentially
  smaller, although we have a new implementation of the oaat hash
  function, alongside the old one, now) and pushb only if needed
  (i.e. state has changed or user has set $RANDOM, but not onfork)
2009-09-23 18:04:58 +00:00
tg 9dd98da40d Support Dave Korn’s alternative 「'a'」 (or 「'…'」) form for base-one
integers in addition to my 「1#a」 (or 「1#…」), which also allows for
finer end-of-character checking. Note that this is locale-dependent in
ksh93, set ±U dependent in mksh, and mksh’s OPTU-16 encoding is used.
2009-09-06 17:55:55 +00:00
tg 9b8d4023fa … but since I liked the bonus of having the hval stored so much, merge
it with the array index; var.c says that
│ 1244         /* The table entry is always [0] */
so that we can have a special flag and a union which stores hval for
the table index, the array index otherwise (coïncidentally *hint hint*
they have the same size)
2009-08-28 21:01:27 +00:00
tg bb7a720a00 for now, until we really use kt*() otherwise, ifdef out tablep/hval 2009-08-28 20:38:43 +00:00
tg 4ccdfc8508 much better solution: save tablep and hash value¹ in the struct tbl entry
① also saves time during texpand :D

XXX this doesn’t work well with the current indexed-array implementation
2009-08-28 20:30:59 +00:00
tg 0e1266ef90 While mksh R39 builds fine on MirOS #7s8E on my trusty sparc, pgcc 2.95.3
throws out quite some warnings – fix most of them except most emitted via
-Wconversion; work around some others; discard bogus warnings.

sync clog
2009-08-08 13:08:53 +00:00
tg 0432f97ffe more KNF, mostly whitespace, this time more manual labour applying style(9) 2009-06-10 18:12:51 +00:00
tg bbf00b5d93 merge fix from oksh:
"let --" was crashing ksh; found by phy0@rambler.ru.  Various other expressions
involving ++ and -- also ran into this. Insufficient checks for end of parse in
the tokenizer made it assume that an lvalue had been found
2009-06-08 20:13:07 +00:00
tg 6b2ad96bac some rather mechanical KNF, now that style(9) clarified on the status
of sizeof() as a function-like unary operator; use parenthesēs around
sizeof and return args consistently too
2009-06-08 20:06:50 +00:00
tg b145ca5c9f • sync distrib/special/mksh/Makefile with bin/mksh/Build.sh and
fix the regression test’s results while here, which have been
  broken since cid 10049D9BE5254CE65B8
• get rid of separate copyright file which was intended for De-
  bian; track down commits in all files of oksh-mirbsd and mksh
  to get correct copyright years per-file, as is BSD custom
2009-05-16 16:59:42 +00:00
tg 8a7223d4cd split utf_ptradjx into utf_ptradj function (to save space) and
self-assignment macro; remove some uses of the macro in favour
of foo += utf_ptradj(foo)
2009-05-16 15:09:07 +00:00
tg 9e83002841 try to do some optimum struct packing except for struct env
(pointers, longs, size_t first; time_t next; int etc. then enum, bool)
2009-04-07 18:41:37 +00:00